In a double-blind multicentre study of parallel group design the efficacy and safety of remoxipride and haloperidol were compared in a total of 96 patients with acute episodes of schizophrenic or schizophreniform disorder according to DSM-III. There were 48 patients in each treatment group; 27 men and 21 women in the remoxipride group, 33 men and 15 women in the haloperidol group. The median duration of illness was 7 years in both groups. The mean daily dose was 437 mg for remoxipride and 10.6 mg for haloperidol during the last week of treatment. No statistically significant differences in total BPRS scores were found between remoxipride and haloperidol. The median total BPRS scores at the start of active treatment were 26 in the remoxipride and 27 in the haloperidol group; these were reduced to 16 and 12.5, respectively, at the last rating. According to Clinical Global Impression (CGI), 43% of patients in the remoxipride group and 68% of those in the haloperidol group improved much or very much during treatment. This difference was not statistically significant. Treatment-emergent extrapyramidal side effects such as akathisia, tremor, and rigidity occurred significantly more frequently in the haloperidol group; this group also made more frequent use of anticholinergic drugs. Neither of the trial drugs seriously affected laboratory or cardiovascular variables. It is concluded that remoxipride has an antipsychotic effect in a dose range of 150-600 mg per day comparable to that of haloperidol in doses up to 20 mg per day but with fewer extrapyramidal side effects.